Combinatorial Selection, Inhibition and Antiviral Activity of DNAThioaptamers Targeting the RNase H Domain of HIV-1 ReverseTranscriptase†

Anoma Somasunderam‡, Monique R. Ferguson§, Daniel R. Rojo§, VaratharasaThiviyanathan‡, Xin Li‡, William A. O’Brien§, and David G. Gorenstein‡,*

‡Department of Human Biological Chemistry & Genetics, University of Texas Medical Branch, Galveston,Texas 77555

§Division of Infectious Diseases, Department of Internal Medicine, University of Texas Medical Branch,Galveston, Texas 77555

AbstractDespite the key role played by the RNase H of Human Immunodeficiency Virus-1 ReverseTranscriptase (HIV-1 RT) in viral proliferation, only a few inhibitors of RNase H have been reported.Using in vitro combinatorial selection methods and the RNase H domain of the HIV RT, we haveselected double-stranded DNA thioaptamers (aptamers with selected thiophosphate backbonesubstitutions) that inhibit RNase H activity and viral replication. The selected thioaptamer sequenceshad a very high proportion of G residues. The consensus sequence for the selected thioaptamersshowed G clusters separated by single residues at the 5′ end of the sequence. Gel electrophoresismobility shift assays and nuclear magnetic resonance spectroscopy showed that the selectedthioaptamer binds to the isolated RNase H domain, but did not bind to a structurally similar RNaseH from E.coli. The lead thioaptamer, R12-2, showed specific binding to HIV-1 RT with a bindingconstant (Kd) of 70 nM. The thioaptamer inhibited the RNase H activity of intact HIV-1 RT. In cellculture, transfection of thioaptamer R12-2 (0.5 μg/ml) markedly reduced viral production andexhibited a dose response of inhibition with R12-2 concentrations ranging from 0.03 μg/ml to 2.0μg/ml (IC50 <100 nM). Inhibition was also seen across a wide range of virus inoculum, ranging frommultiplicity of infection (m.o.i) of 0.0005 to 0.05, with reduction of virus production by more than50% at high m.o.i. Suppression of virus was comparable to that seen with AZT at m.o.i. ≦ 0.005.

Reverse transcription of viral genomic RNA into DNA, an essential step in the replication ofHuman Immunodeficiency Virus type 1 (HIV-1)1 and other retroviruses, is catalyzed byreverse transcriptase. The bifunctional HIV-1 RT is a heterodimer with 66 and 51 kDa subunits(1). Both subunits contain a DNA polymerase domain. The 66 kDa domain, p66, contains anRNase H domain in addition to the DNA polymerase domain. The DNA polymerase utilizesboth RNA and DNA templates to accomplish viral genomic replication. RNase H catalyzescleavage of the viral RNA strand from the DNA:RNA hybrid duplex, thereby releasing theviral DNA copy to ultimately integrate into the host cell’s genome.

HIV infection progresses to AIDS at variable rates in virtually every infected individual overa period of several years in the absence of therapy. RT inhibitors were the first class of

†This work was supported by grants from the NIAID / National Institute of Health (U01-A1054827 and A127744 to D. G. G and R21AI058194-01A2 to M. R. F.), and the Welch Foundation (H-1296, to D.G.G.).*Corresponding author: David G. Gorenstein, Department of HBC&G, 301 University Blvd., Galveston, TX 77555-1157. Phone: 409747 6800. Fax: 409 747 6850. E-mail: david@nmr.utmb.edu

NIH Public AccessAuthor ManuscriptBiochemistry. Author manuscript; available in PMC 2008 September 8.

Published in final edited form as:Biochemistry. 2005 August 2; 44(30): 10388–10395. doi:10.1021/bi0507074.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

antiretroviral drugs approved and used to treat HIV infection. Although newer drugs targetingthe HIV protease and, more recently, the envelope proteins, have been approved for use, RTinhibitors remain an important component of combination therapy known as highly active anti-retroviral therapy (HAART). Most commonly used HAART regimens typically include twoor three RT inhibitors. The use of HAART, rather than less intensive antiretroviral regimens,has dramatically reduced the rates of progression to AIDS and has improved survival (2).Currently available RT inhibitors target the polymerase activity and are either nucleosideanalog drugs that bind at the polymerase active site or are non-nucleoside inhibitors that bindto a different region of the HIV-1 RT. Therapeutic benefits can be markedly diminished by theemergence of drug resistant strains and by the potential toxicity of the approved antiretroviraldrugs (3,4). Resistance to these drugs emerges rapidly because of the highly error-prone RT,and by the potential for recombination between different strains when cells are co-infected(5). Due to extensive cross-resistance within each drug class that reduces antiviral activity andclinical benefit of sequential HAART, there is an urgent need to develop new agents andtreatment approaches to fight AIDS. This search for alternate targets and new agents has ledto a new class of RT inhibitors known as Oligonucleotide Reverse Transcriptase Inhibitors(ONRTIs) (6,7).

ODNs are emerging as potential therapeutic agents that can be attractive alternatives for thechemical drugs (8,9). ODNs operate via different mechanisms, such as the sequence-specifictranslational arrest of mRNA expression (antisense), RNA inhibition using specific shortdouble-stranded RNA (siRNA), or by binding to and inhibiting RT or other HIV proteins/cellular receptors (aptamers or decoys) (10-12). The antisense strategy, proposed for thepotential treatment of several diseases including AIDS and cancer (13), is based on the bindingof ODNs to the mRNA by complementary base pairing. The decoy (or sense) strategy relieson selective binding of the ODNs, possibly selected as aptamers, to a nucleic acid bindingprotein. The decoy aptamer mimics the protein’s natural ligand and, therefore, competes withit for complex formation with the protein. This association traps the protein in a nonfunctionalcomplex. An advantage of aptamers over the antisense technology is that the aptamer ODNs

1The abbreviations used are:

EMSA electrophoretic mobility shift assay

HAART highly active anti-retroviral therapy

HIV-1 human immunodeficiency virus type 1

HSQC heteronuclear single quantum coherence

m.o.i. multiplicity of infection

NMR nuclear magnetic resonance

NOESY nuclear Overhauser enhancement spectroscopy

ODN oligonucleotide agents

RNase ribonuclease

RT reverse transcriptase

Somasunderam et al. Page 2

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

do not have to overcome the thermodynamic penalty required for the unfolding of their ownand the target’s secondary and tertiary structures. RNA and DNA aptamers targeting severalHIV proteins have been reported. This includes the inhibition of HIV-1 RT’s polymerase(14,15) and RNase H activities (16,17), HIV-1 integrase function (18), and the capsid proteinNCp7 (19,20).

Backbone modifications in ODNs are essential to increase their resistance toward digestion bycellular nucleases. Sulfur substitutions of phosphoric oxygens of the DNA backbone, as inmonothiophosphate and dithiophosphate aptamers, often enhance their binding to proteins(21,22). One can also take advantage of the increased affinity of the ODNs generally associatedwith sulfur substitutions, however, complete substitution of the backbone could lead to non-specific binding and therefore to a loss in specificity of the ODN to the target protein. Therefore,the number and position of sulfur substitutions in these thioaptamers need to be optimized todecrease binding to non-target proteins while enhancing binding to the target protein.

Despite the key role played by HIV RT RNase H in viral proliferation, aptamers thatspecifically target this domain have not yet been reported. Although aptamers inhibiting RNaseH activity have been reported, they were selected against the entire HIV RT. We have used theRNase H domain of HIV-1 RT in our selection process to select, by definition, thioaptamersthat bind directly to RNase H. In this paper, we report the in-vitro combinatorial selection ofmonothio-phosphate modified DNA thioaptamers that selectively bind to the RNase H domainof HIV-1 RT, inhibit RNase H activity, and exhibit antiretroviral activity.

MATERIALS AND METHODSMaterials

NTPs and HIV-1 RT were purchased from Amersham Bioscience (Piscataway, NJ). 3H-UTPwas purchased from NEN (Boston, MA). The chirally pure Sp isomer of dATP (αS) wasobtained from Biolog Life Science Institute (Bremen, Germany). TOPO TA cloning kits wereobtained from Invitrogen (Carlsbad, CA), the plasmid isolation kits were obtained from Qiagen(Foster City, CA), and AmpliTaq DNA polymerase and other PCR consumables werepurchased from Applied Biosystems (Foster City, CA). The transfection agent Oligofectamine(OF) was purchased from Invitrogen.

Preparation of HIV-1 RNase HWe have constructed a new T7 RNA polymerase/IPTG-inducible plasmid whose insert encodesthe 15 kDa RNase H domain of HIV-1 RT. The plasmid containing the coding region of HIV-1RNase H (strain HXB2) was cloned into pET30a(+) vector (Novagen), transformed andexpressed E.coli, and purified as described previously (23). Both unlabeled anduniformly 15N-labeled protein were purified to homogeneity after the over expression inminimal media or minimal media containing 15NH4Cl. This procedure yields excellentquantities of pure RNase H (ca. 20 mg/L in an E. coli expression system). Expression tests inrich, minimal and 15N-labeled minimal media provided the 15 kDa fragment (by SDS PAGE).The authenticity of the protein was confirmed by N-terminal sequencing and NMRspectroscopy.

Synthesis of DNA LibrariesThe chemically synthesized random combinatorial library is a 62 nucleotide long single strandDNA containing a 22 nucleotide random region flanked by 19- and 21-nucleotide long PCRprimer regions (Figure 1). The library was annealed with the reverse primer, subjected toKlenow reaction for 5 hrs at 37 °C, and amplified by PCR using AmpliTaq DNA polymeraseand a mixture of dATP(αS), dTTP, dCTP, and dGTP to give the thioaptamer-substituted library.

Somasunderam et al. Page 3

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Reaction conditions for the PCR-amplifications were: oligonucleotide library (40 nM), dATP(αS) (160 μM), mixture of dTTP, dCTP, and dGTP (80 μM each), MgCl2 (2 mM), primer (400nM of each) and AmpliTaq DNA polymerase (1 U) in a total volume of 100 μl. PCR was runfor 35 cycles of 94 °C (2 min), 55 °C (2 min) and 72 °C (2 min). The resulting 62-mer librarycontained monothiophosphate substitutions at the phosphate 5′ to every dA residue, in the Spconfiguration with the exception of the primer region on the non-template strand. Standardphosphoryl PCR amplification was carried out with a mixture of dNTPs (80 μM each) alongwith the other reagents for 25 cycles of 94 °C (1 min), 45 °C (1 min) and 72 °C (1 min). Forthe EMSA, 5′-fluorescein labeled thioaptamers were synthesized using a PCR primer labeledwith fluorescein at the 5′- end. We observed incomplete amplification of the thioaptamersresulting in shorter sequences. This is caused by the formation of secondary structuresparticularly in GC rich regions (24). To minimized this formation of secondary structures ofthe DNA template we used Betaine solution (1M, Sigma) and DMSO (5%v/v, Sigma) in thePCR reaction mixture.

Combinatorial Selection of ThioaptamersPurified RNase H protein was incubated with the previously nitrocellulose-filtered (to excludesequences that bind to nitrocellulose) PCR-amplified, monothio DNA library in a bindingbuffer containing 50 mM Tris-HCl, 1.25 mM MgCl2, 25-400 mM KCl and 10 mM DTT in atotal volume of 50 μl for 1 hr at ambient temperature, then passed over nitrocellulose filters.The filters were washed (4 × 1ml) with binding buffer to remove the unbound and weakly-bound DNA. After the washes, thioaptamers bound strongly to the protein are retained on thefilter. A 0.5 ml solution of 8 M urea and 2 M KCl was added to elute the thioaptamers boundto the protein. A negative control experiment without the protein was performedsimultaneously to monitor any nonspecific binding of the thiophosphate library to the filter.The eluted DNA was extracted and PCR amplified for use in the next round of selection. Theamplified DNA was analyzed by non-denaturing 15% polyacrylamide gel electrophoresis. Thestringency of selection was tightened at each selection round by decreasing the amount ofprotein and by gradually increasing the salt concentration (from 25 to 400 mM) in the bindingbuffer. A subset of sequences from the initial library and from selection rounds 5, 8, 12 and14, were determined. A chemically synthesized double stranded 62-mer, 5′ATAGTTACTAACTGACAGTACACTAGTGAACTTGATGTTAGACTGAAGTACTTTAACATACC3′, was used as a control for the inhibition experiments. Thesequence of this random 62-mer does not have any transcription factor binding sites. Since NF-κB and AP-1 transcription factor binding sites are found in the promotor region of the HIV-1LTR (25,26), ODNs able to bind to these sites would interfere with viral replication. We alsotested the effect of an AP-1 specific ODN, XBY-S2. Thioaptamer XBY-S2 (27) is a 14-merwith 6 dithio-phosphate substitutions (5′CCAGTGACTCAGTG-3′) that includes a consensusAP-1 binding site.

Electrophoresis Mobility Shift AssayOne of the selected thioaptamers, R12-2, was used to test the specific binding to RNase H.R12-2 was enzymatically labeled at the 5′-end with fluorescein and was incubated withincreasing concentrations of the protein (isolated RNase H domain or intact HIV-1 RT RNaseH) in 50 mM Tris-HCl buffer (pH 8), 10 mM MgCl2 and 40 mM DTT at 37 °C for 30 minutes.The reaction mixture was loaded onto a native 8% polyacrylamide gel, which waselectrophoresed and analyzed on a FluorChem 8800 imager (Alpha Innotech, CA). A similarexperiment was performed with the E.coli RNase H, to test whether R12-2 binds to the RNaseH from E.coli.

Somasunderam et al. Page 4

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

HIV-1 RT RNase H Activity AssayA DNA: RNA heteroduplex containing radiolabeled RNA was prepared as a substrate forRNase H. A tritium-labeled 18-mer RNA molecule, the stem loop 2 of the HIV-1 psi-RNA(28), was made by in-vitro transcription using T7 RNA polymerase with a mixture of ATP,CTP, GTP and 3H-UTP. Transcribed RNA was purified by centrifugation using MicroconYM-3 filters (Amicon) to remove unused NTPs and short RNA sequences. The purified RNAwas annealed with the complementary DNA strand to form an RNA:DNA hybrid in which theRNA strand was 3H-labeled. HIV-1 RT (0.35 μM) was incubated with R12-2 thioaptamer (0.5to 2.0 μM) in a buffer containing 50 mM Tris-HCl (pH 8.0), 6 mM MgCl2, 10 mM DTT and80 mM KCl for 30 min at 37 °C. To this mixture, labeled RNA:DNA hybrid (40 000 cpm) wasadded to a final volume of 50 μl and incubated at 37 °C for 20 min. The reaction was quenchedby the addition of 150 μl of 10% trichloroacetic acid, and the mixture was placed on ice for 10min. The reaction mixture was filtered and washed, and the amount of radioactivity in thefiltrate was measured using a scintillation counter (Beckman Coulter, CA). Radioactivity inthe filtrate would be proportional to RNase H activity. A control experiment was performedwithout any HIV-1 RT, and 100% RNase H activity is assumed in the absence of any R12-2thioaptamer.

NMR SpectroscopyUniformly 15N-labeled RNase H and chemically synthesized R12-2 were used to monitor thebinding of the R12-2 thioaptamer to the isolated RNase H domain. The oligonucleotideconcentrations were calculated from the absorbance values and the molar absorptioncoefficients calculated using the nearest neighbor parameters. 15N-HSQC spectra (29) of theisolated RNase H in the presence of 25 mM MgCl2 were acquired on a Varian UnityPlus 750MHz instrument equipped with pulse field gradients and triple resonance probes. Onedimensional proton NMR and 2D NOESY spectra (250 ms mixing time) of R12-2 thioaptamerin 95% H2O/ 5% D2O were collected. The signals from the R12-2 imino protons, which arevery sensitive to changes in the DNA conformation, were monitored during the titration ofRNase H to the R12-2.

Viruses and CellsHIV-1 SF162-R5 is a primary isolate that uses CCR5 as a co receptor (R5), and was obtainedfrom the NIH AIDS Research and Reference Reagent Program. U373-MAGI-CCR5 cells havemodifications of the U373 astrocytoma adherent cells that are used for HIV transfection andinfection experiments. U373-MAGI-CCR5 cells express β-gal under the control of HIV LTR,which is trans-activated by the HIV Tat protein in relation to the level of virus replication. Inaddition, these cells express CD4 and the human chemokine receptor CCR5 on its surface,which allow infection by primary HIV-1 strains. U373 cells are propagated in 90% Dulbecco’sModified Eagle Medium (DMEM), 10% fetal bovine serum, 0.2 mg/ml G418, 0.1 mg/mlhydromycin B, and 1.0 μg/ml puromycin. For transfection and infection experiments, U373cells were maintained in 90% DMEM, 10% fetal bovine serum, and 1% penicillin/streptomycin. Thioaptamer R12-2, AP-1-specific XBY-S2 or the random 62-mer ODNs weretransfected into the cells 24 hr prior to infection using Oligofectamine (OF) liposomes,according to the manufacturer’s instructions (Invitrogen).

Infection of ODN-Transfected U373-MAGI-CCR5 CellsTransfected cells and control cells were infected with HIV-1 at m.o.i. of 0.003 (whichconsistently gives high levels of viral production) for 2 hr. Equal volumes of medium werethen added and the infection was continued for 48 hr. HIV-1 replication was quantified in celllysates by measurement of β-gal activity, as determined by luminometry using the β-Glo kit(Promega) according to the manufacturer’s guidelines. β-gal activity is expressed as relative

Somasunderam et al. Page 5

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

light units (RLU). To determine HIV-1 TCID50 (50% tissue culture infective dose) in U373-MAGI-CCR5 cells, serial dilutions (10x) of SF162-R5 virus supernatant were inoculated ontoa 24-well plate containing 3×104 U373 cells for 2 hr, then equal volumes of complete mediumwere added, and the infection was continued for 48 hr. Cells then were lysed and assayed forHIV-directed β-gal expression, and TCID50 values were calculated from quadruplicates. Thep24 antigen concentrations were measured (p24 antigen capture assay, Beckman-Coulter) todetermine virus production in culture supernatants harvested at day 7 post-infection.

RESULTSPCR Amplification of Thioaptamers

For the amplification of DNA sequences with monothio phosphate substitutions, dATP(αS)was used in the PCR reaction mixture (thio-PCR) along with dGTP, dTTP and dCTP. Onlythe Sp isomer of the dNTP(αS) is used as a substrate by the AmpiliTaq DNA Polymeraseenzyme to yield the pure Rp stereoisomer (30). Yields from the thio-PCR were 20-35% lowercompared to yields obtained using all four normal dNTPs. To compensate for the loss of yieldin the thio-PCR, reaction conditions were changed by doubling the concentrations of dATP(αS), increasing the annealing temperature by 10 °C from 45 °C to 55 °C and by increasingthe number of PCR cycles from 25 to 35.

Sequences of Selected ThioaptamersAfter twelve rounds of selection, the monothio library converged to a few predominantsequences out of the starting library containing (hypothetically) 1014 different sequences.Sequences from the variable region of the thioaptamers selected after the 12th and 14th roundsshown in Figure 2, were grouped into three classes based on primary sequence alignment bythe Clustal W program (31). Sequences that showed very high degrees of alignment are placedin class I, while those that do not align well are listed under class II. The aptamers with truncatedvariable regions are listed under class III. These truncated thioaptamers had 10-15 base pairsin the variable region as opposed to 22 base pairs in the original library. It is possible that theshorter thioaptamers (Class III) could have been formed by inefficient or incompleteamplification of the library during the PCR cycles. However, sequences in all three classesshowed similar sequence motifs. The sequences of the selected thioaptamers showed G-richsequences at their 5′ ends. These G rich sequences, with a minimum of four interspersed Gnucleotides, are known to have the potential to form G-quartet structures (32,33,34). Recentreports of single strand DNA aptamers selected against HIV-1 RT, HIV-1 integrase, and humanRNase H also showed G-rich sequences that could form G-quartet structures (16,17,18,35).

EMSAs Showed Selective Binding of R12-2 to the RNase H Domain of HIV-1 RTIn EMSA experiments, the lead thioaptamer R12-2 showed specific binding to the isolatedRNase H domain of HIV-1 RT (Figure 3). The intensity of the free DNA band (lower band)decreased with increasing RNase H concentration, showing the increased binding of R12-2with the increase in protein concentration. Under identical conditions, however, the R12-2 didnot show any significant binding to the RNase H of E.coli, a structurally similar protein.Similarly, the initial library did not show any binding to the isolated RNase H domain of HIV-1RT nor to the intact HIV-1 RT (Figure 3).We also tested the binding of R12-2 to the intactHIV-1 RT. Two bands were visible in these gels (Figure 4A). The lower band corresponds tothe free oligonucleotide, and the shifted upper band corresponds to the oligonucleotide boundto the protein. Presence of protein in the upper band was detected by staining the gel withCoomasie blue. With increasing protein concentrations, the intensity of the protein-bound DNAband (upper band) increased while the intensity of the free DNA band (lower band) decreased.This observation shows that R12-2 is binding to the HIV-1 RT. The quantitative binding ofR12-2 to the HIV-1 RT is shown in the plot of bound band intensity vs. protein concentration

Somasunderam et al. Page 6

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

(Figure 4B). The dissociation constant (Kd), determined from non-linear regression analysiswith the Hill Plot method using the Origin software was 70 nM.

NMR Spectroscopy Showed R12-2 Interacts with the RNase H Domain of HIV-1 RTWe used NMR spectroscopic methods to demonstrate the binding of R12-2 to the isolatedRNase H domain (data not shown). Imino proton signals of R12-2, the most sensitive indicatorsof the structural perturbations in DNA conformation, showed broadening with the addition ofRNase H, indicating the binding of R12-2 to the protein. The line widths of the imino protonsignals increased with the increase in protein concentration and these signals eventuallydisappeared.

Inhibition of RNase H Activity by Thioaptamer R12-2The ability of R12-2 to inhibit the RNase H activity was tested using a radio-labeled RNA:DNAhybrid duplex, as described in the methods. Although the RNase H domain can be expressedand isolated as a stable protein, the isolated protein is inactive or very weakly active (23,36,37). Therefore, for the inhibition assay (Figure 5), we used the intact HIV-1 RT instead of theisolated RNase H domain. The assay was based on the ability of the RNase H catalytic functionin intact HIV-1 RT to cleave the RNA strand in the RNA:DNA hetero-duplex. Addition of 0.5μM of R12-2 decreased the RNase H activity to 47% compared to the control, where no R12-2was present. Additional R12-2 decreased the RNase H activity further down to 39%. However,when the R12-2 concentration was increased higher than 2 μM, no further reduction in RNaseH activity was observed.

R12-2 showed dose-dependent inhibition of HIV-1 replication in vitroIncubation of U373-MAGI-CCR5 cells, which had been transfected with Oligofectamine (OF)and thioaptamer R12-2 24 h prior to infection with HIV-SF162 (R5; 2 ng, at m.o.i. of 0.003),a prototypic primary HIV strain, resulted in significant reduction of infection, as compared tothe cells transfected with the random 62-mer (Figure 6). Inhibition of the HIV-1 infection byAZT is shown for comparison. The ODN specific to the human AP-1, XBY-S2, also showedmodest effects on HIV infection (Figure 7). However, the inhibition by the R12-2 issignificantly stronger than the inhibition by XBY-S2. The sequence of XBY-S2 includes aconsensus AP-1 binding site, and would influence the expression of immune responses in thecell by sequestering AP-1 proteins. The modest inhibition exhibited by XBY-S2 may be dueto its ability to serve as a decoy to bind to AP-1 transcription factors. A dose-dependentinhibitory response was demonstrated, with R12-2 doses ranging from 0.03 to 2.0 μg/ml(IC50 < 100 nM) to the same low level as 10 μM AZT (Figure 8). However, the AP-1 specificthioaptamer XBY-S2 did not show significant change in inhibition up to the dose level of 0.25μg/ml. At higher concentrations (> 0.5 μg/ml) both the XBY-S2 and R12-2 did not exhibitsignificant change in the percentage of inhibition. Transfection of R12-2 (0.05 μg/ml) resultedin the inhibition of viral replication across a broad range of SF162 (R5) HIV-1 inocula rangingfrom an m.o.i. of 0.0005 (0.3 ng) to 0.05 (30 ng) as shown in Figure 9. R12-2 showed significantinhibition of the virus at high m.o.i., and viral replication is comparable to that seen with AZTat m.o.i. ≤ 0.005 (0.3 ng p24 as measured by the p24 antigen capture ELISA method). R12-2does not show any cytotoxicity to the cells (data not shown).

DISCUSSIONODNs are emerging as promising alternative therapeutic agents. However, to increase theresistance to digestion by cellular enzymes, modified ODNs have to be developed.Modifications in the phosphate backbone will significantly affect their binding to targetproteins, because most of the direct contacts between DNA-binding proteins and their bindingsites in DNA involved the phosphate groups (38,39). Sulfur substitutions of the phosphoric

Somasunderam et al. Page 7

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

oxygens of ODNs often lead to their enhanced binding to numerous proteins (21,22). However,complete substitution of phosphoryl oxygens with sulfur appears to make thioaptamers too“sticky” such that they lose their ability to discriminate between target and non-target proteins.Therefore, the total number and positions of thio-substituted phosphates have to be optimizedusing either rational design principles or by combinatorial techniques to decrease non-specificprotein binding, and to enhance only the specific favorable interactions with proteins such asRNase H. Phosphorothioate modifications show increased resistance to digestion by cellularnucleases and often have higher affinity for binding to proteins. They can also be readilysynthesized in high yields and enhance favorable pharmacokinetic properties (40). Variouspharmacokinetic studies in animals have shown that phosphorothioate analogs are rapidlydispersed to various tissues and cleared through the kidney within acceptable time periods(41). By using only the dATP (αS) during PCR amplification cycles, we have producedthioaptamers that have thio-substitutions only at positions 5′ to the dA residues. Combinatorialselection methods that allow screening of a large number of random sequences in nucleic acidlibraries for affinity to proteins facilitate the selection of ODNs that would have the optimalnumber of thiophosphate substitutions. Our lead thioaptamer, R12-2, showed selective bindingto the RNase H of HIV-1 RT. This thioaptamer did not bind to the E. coli RNase H, even thoughthese two proteins are structurally similar. By using selection methods and combinatorialsynthesis which incorporates selective sulfur substitution, we have selected a DNA thioaptamerthat binds tightly and specifically to the RNase H domain of HIV-1 RT, inhibits the RNase Hactivity in-vitro and inhibits HIV-1 replication in cell cultures. The selected thioaptamer didnot bind to the E.coli RNase H, structurally similar protein

Although the RNase H domain of the HIV-1 RT can be expressed and isolated as a stableprotein, the isolated domain is either catalytically inactive or very weakly active. This loss ofcatalytic activity in the isolated domain is probably not caused by structural differencesbetween the two forms, because the RNase H structures in isolation and in the intact RT arevery similar (42,43). The basis for the inactivity of the isolated domain is attributed mainly tothe increased dynamics. Recent NMR studies on the backbone dynamics of the isolated RNaseH domain indicate that the intramolecular dynamic behavior of the isolated domain is severe,resulting in the loss of catalytic activity of RNase H in isolation (44,45). Since we have usedthe isolated RNase H domain in our iterative selection rounds, the selected thioaptamers arepresumed to specifically bind to the RNase H domain of intact RT. However, the R12-2 is a62-mer duplex and other regions of the duplex very likely extend into the polymerase site ofthe RT, as observed in various duplex DNA-RT X-ray crystal structures and models (46,47).The R12-2 reduced the RNase H activity of RT by 50%. However, the inhibition of viralproduction was more dramatic (Figure 6). It is quite possible that the difference in the inhibitionlevels between the RNase H activity and the viral production could be the result of R12-2′sinhibition of the polymerase activity of RT. The selected thioaptamer, R12-2, inhibits thefunction of the RNase H in the intact HIV-1 RT. Because the structures of the RNase H domainare similar in isolation and in intact HIV-1 RT, selecting thioaptamers against the isolateddomain that could inhibit the function of the intact HIV-1 RT would be a better choice thanselecting aptamers against the intact HIV-1 RT. Selecting against intact HIV-1 RT does notguarantee that the selected aptamer will bind to the RNase H domain. In fact, the selectedaptamer may bind to the polymerase domain rather than to the RNase H domain. Perhapssequences selected against RNase H domain and intact HIV-1 RT could be used adjunctively.As far as we are aware, this is the first report of aptamer that was selected to target the isolateddomain of RNase H.

The R12-2 exhibited significant, dose-dependent inhibition of HIV-1 infection. The AP-1specific thioaptamer XBY-S2 also showed modest inhibition of infection at higher aptamerconcentrations (> 0.5 μg/ml). The XBY-S2 thioaptamer was found to bind AP-1 proteins froma macrophage like cell line, and inhibits the binding of AP-1 transcription factor to DNA. (D.

Somasunderam et al. Page 8

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

G. Gorenstein, personal communication). An AP-1 binding site is shown to be present in thepromotor region of the HIV-1 LTR (25). The modest inhibition of HIV-1 infection by XBY-S2 might be due to its ability to bind to AP-1. Inhibition by R12-2 was seen across a broadrange of HIV inocula, and it was comparable to AZT treatment at low m.o.i. The inhibition athigh m.o.i is noteworthy, since transfection efficiency (“uptake”) of the virus or thioaptamerin some experiments was only 80%. Thus, the thioaptamer activity may be underestimatedwhen compared with the activity of drugs, such as AZT, which are taken up by essentiallyevery viable cell in the culture.

As the HIV epidemic continues to mature through treatment error and emerging resistance tothe initial classes of antiretroviral drugs, which include the HIV-1 RT inhibitors, there will bea need to develop treatments directed to new targets. The various inhibitors of HIV entry,targeting attachment, co-receptor binding and fusion will have a substantial impact on theability to suppress viremia as they are developed for clinical use by extending treatment options,but, additional new classes of drug will inevitably be needed. Combinatorial selection ofthioaptamers that bind to novel targets may allow development of therapeutic agents directedto diverse viral functions.

ACKNOWLEDGEMENTWe thank Dr. R. Hodge and the NIEHS synthetic core facility at the UTMB for the chemical synthesis of DNAs, Dr.X. Yang for providing the XBY-S2 sample and Drs. N. Herzog, D. Kallick, and D. Volk for helpful discussions.

REFERENCES1. Schatz O, Mous J, Le Grice SFJ. HIV-1 RT-associated ribonuclease H displays both endonuclease and

3′----5′ exonuclease activity. EMBO J 1990;9:1171–1176. [PubMed: 1691093]2. Palella FJ Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, Aschman DJ, Holmberg

SD, HIV Outpatient Study Investigators. Declining morbidity and mortality among patients withadvanced human immunodeficiency virus infection. N. Engl. J. Med 1998;338:853–860. [PubMed:9516219]

3. Yeni PG, Hammer SM, Hirsch MS, Saag MS, Schechter M, Carpenter CC, Fisch,l MA, Gatell JM,Gazzard BG, Jacobsen DM, Katzenstein DA, Montaner JS, Richman DD, Schooley RT, ThompsonMA, Vella S, Volberding PA. Treatment for adult HIV infection: 2004 recommendations of theInternational AIDS Society-USA Panel. J. A .M. A 2004;292:251–265.

4. Hirsch MS, Brun-Vezinet F, Clotet B, Conway B, Kuritzkes DR, D’Aquila RT, Demeter LM, HammerSM, Johnson VA, Loveday C, Mellors JW, Jacobsen DM, Richman DD. Antiretroviral drug resistancetesting in adults infected with human immunodeficiency virus type 1: 2003 recommendations of anInternational AIDS Society-USA Panel. Clinical Infectious Diseases 2003;37:113–128. [PubMed:12830416]

5. Ferguson MR, Rojo DR, von Lindern JJ, O’Brien WA. HIV-1 replication cycle. Clin. Lab. Med2002;3:611–635. [PubMed: 12244589]

6. Majumdar C, Stein CA, Cohen JS, Broder S, Wilson SH. Stepwise mechanism of HIV reversetranscriptase: primer function of phosphorothioate oligodeoxynucleotide. Biochemistry1989;28:1340–1346. [PubMed: 2469468]

7. Tuerk C, MacDougal S, Gold L. RNA pseudoknots that inhibit human immunodeficiency virus type1 reverse transcriptase. Proc. Natl. Acad. Sci U.S.A 1992;89:6988–6992. [PubMed: 1379730]

8. Stein CA, Cheng YC. Antisense oligonucleotides as therapeutic agents-is the bullet really magical?Science 1993;261:1004–1012. [PubMed: 8351515]

9. Morgan RA, Walker R. Gene therapy for AIDS using retroviral mediated gene transfer to deliver HIV-1antisense TAR and transdominant Rev protein genes to syngeneic lymphocytes in HIV-1 infectedidentical twins. Hum. Gene Ther 1996;7:1281–1306. [PubMed: 8793552]

10. Ellington AD, Szosytak JW. In vitro selection of RNA molecules that bind specific ligands. Nature1990;346:818–822. [PubMed: 1697402]

Somasunderam et al. Page 9

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

11. Gold L, Polisky B, Uhlenbeck OC, Yarus M. Diversity of oligonucleotide functions. Annu. Rev.Biochem 1995;64:763–797. [PubMed: 7574500]

12. King DJ, Bassett SE, Li X, Fennewald SA, Herzog NK, Luxon BA, Shope R, Gorenstein DG.Combinatorial selection and binding of phosphorothioate aptamers targeting human NF-kappa BRelA(p65) and p50. Biochemistry 2002;41:9696–9706. [PubMed: 12135392]

13. Agarwal, S. Antisense oligonucleotides: possible approaches for the chemotherapy of AIDS. In:Wickstrom, E., editor. Prospects for antisense nucleic acid therapy of cancer and AIDS. Liss; NewYork, N.Y: 1991. p. 143-159.

14. Schneider DJ, Feigon J, Hostomsky Z, Gold L. High-affinity ssDNA inhibitors of the reversetranscriptase of type 1 human immunodeficiency virus. Biochemistry 1995;34:9599–9610. [PubMed:7542922]

15. Nickens DG, Patterson JT, Burke DH. Inhibition of HIV-1 reverse transcriptase by RNA aptamersin Escherichia coli. RNA 2003;9:1029–1033. [PubMed: 12923252]

16. Andreola ML, Pileur F, Calmels C, Ventura M, Tarrago-Litvak L, Toulme JJ, Litvak S. DNA aptamersselected against the HIV-1 RNase H display in vitro antiviral activity. Biochemistry 2001;40:10087–10094. [PubMed: 11513587]

17. Pileur F, Andreola ML, Dausse E, Michel J, Moreau S, Yamada H, Gaidamakov SA, Crouch RJ,Toulme JJ, Cazenave C. Selective inhibitory DNA aptamers of the human RNase H1. Nucleic AcidsRes 2003;31:5776–5788. [PubMed: 14500841]

18. de Solutrait VR, Lozach P, Altmeyer R, Tarrago-Litvak L, Litvak S, Andreola ML. DNA aptamersderived from HIV-1 RNase H inhibitors are strong anti-integrase agents. J. Mol. Biol 2002;324:195–203. [PubMed: 12441099]

19. Paoletti AC, Shubsda MF, Hudson BS, Borer PN. Affinities of the nucleocapsid protein for variantsof SL3 RNA in HIV-1. Biochemistry 2002;41:15423–15428. [PubMed: 12484783]

20. Shubsda MF, Paoletti AC, Hudson BS, Borer PN. Affinities of packaging domain loops in HIV-1RNA for the nucleocapsid protein. Biochemistry 2002;4:5276–5282. [PubMed: 11955077]

21. Milligan JF, Uhlenbeck OC. Determination of RNA-protein contacts using thiophosphatesubstitutions. Biochemistry 1989;28:2849–2855. [PubMed: 2663062]

22. Marshall WS, Caruthers MH. Phosphorodithioate DNA as a potential therapeutic drug. Science1993;259:1564–1570. [PubMed: 7681216]

23. Becerra SP, Clore GM, Gronenborn AM, Karlstrom AR, Stahl SJ, Wilson SH, Wingfield PT.Purification and characterization of the RNase H domain of HIV-1 reverse transcriptase expressedin recombinant Escherichia coli. FEBS Letters 1990;270:76–80. [PubMed: 1699794]

24. Henke W, Herdel K, Jung K, Schnorr D, Loening SA. Betaine improves the PCR amplification ofGC-rich DNA sequences. Nucleic Acids Res 1997;25:3957–3958. [PubMed: 9380524]

25. Liou HC, Baltimore D. Regulation of the NF-αB/rel Transcription Factor and I αB Inhibitor System.Curr.Opin.Cell Biol 1993;5:477–487. [PubMed: 8352966]

26. Nabel G, Baltimore D. An inducible transcription factor activates expression of humanimmunodeficiency virus T cells. Nature 1987;326:711–713. [PubMed: 3031512]

27. Yang X, Gorenstein DG. Progress in thioaptamer development. Current Drug Targets 2004;5:705–715. [PubMed: 15578951]

28. Amarasinghe GK, De Guzman RN, Turner RB, Chancellor KJ, Wu ZR, Summers MF. NMR structureof the HIV-1 nucleocapsid protein bound to stem-loop SL2 of the psi-RNA packaging signal.Implications for genome recognition. J. Mol. Biol 2000;301:491–511. [PubMed: 10926523]

29. Bax A, Pochapsky SS. Optimized recording of heteronuclear multidimensional NMR spectra usingpulsed field gradients. J. Magn. Reson 1992;99:638–643.

30. Eckstein F, Thomson JB. Phosphate analogs for study of DNA polymerases. Methods Enzymology1995;262:189–202.

31. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressivemultiple sequence alignment through sequence weighting, position-specific gap penalties and weightmatrix choice. Nucleic Acids Res 1994;22:4673–4680. [PubMed: 7984417]

32. Williamson JR, Raguraman MK, Cech TR. Monovalent cation-induced structure of telomeric DNA:the G-quartet model. Cell 1989;59:871–880. [PubMed: 2590943]

Somasunderam et al. Page 10

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

33. Smith FW, Feigon J. Quadruplex structure of Oxytricha telomeric DNA oligonucleotides. Nature1992;356:164–168. [PubMed: 1545871]

34. Bianchi A, de Lange T. Ku binds telomeric DNA in vitro. J. Biol. Chem 1999;274:21223–21227.[PubMed: 10409678]

35. Phan AT, Kuryavyi V, Ma JB, Faure A, Andreola ML, Patel DJ. An interlocked dimeric parallel-stranded DNA quadruplex: a potent inhibitor of HIV-1 integrase. Proc. Natl. Acad. Sci. U.S.A2005;102:634–639. [PubMed: 15637158]

36. Evans DB, Brawn K, Deibel MR Jr. Tarpley WG, Sharma SK. A recombinant ribonuclease H domainof HIV-1 reverse transcriptase that is enzymatically active. J. Biol. Chem 1991;266:20583–20585.[PubMed: 1718968]

37. Hostomsky Z, Hostomska Z, Hudson GO, Moomaw EW, Nodes BR. Reconstitution in vitro of RNaseH activity by using purified N-terminal and C-terminal domains of human immunodeficiency virustype 1 reverse transcriptase. Proc. Natl. Acad. Sci. U.S.A 1991;88:1148–1152. [PubMed: 1705027]

38. Botuyan MV, Keire DA, Kroen C, Gorenstein DG. 31P nuclear magnetic resonance spectra anddissociation constants of lac repressor headpiece·duplex operator complexes: the importance ofphosphate backbone flexibility in protein·DNA recognition. Biochemistry 1993;32:6863–6874.[PubMed: 8334119]

39. Cho Y, Zhu FC, Luxon BA, Gorenstein DG. 2D 1H and 31P NMR spectra and distorted A-DNA-likeduplex structure of a phosphorodithioate oligonucleotide. J. Biomol. Struct. Dyn 1993;11:685–702.[PubMed: 8129879]

40. Tonkinson JL, Guvakova M, Khaled Z, Lee J, Yakubov L, Marshall WS, Caruthers MH, Stein CA.Cellular pharmacology and protein binding of phosphoromono-thioate and phosphorodithioateoligodeoxynucleotides: a comparative study. Antisense Res. Dev 1994;4:269–278. [PubMed:7537561]

41. Gallo M, Montserrat JM, Iribarren AM. Design and applications of modified oligonucleotides. Braz.J. Med. Biol Res 2003;36:143–151. [PubMed: 12563516]

42. Davies JF, Hostomska Z, Hostomsky Z, Jordan SR, Matthews DA. Crystal structure of theribonuclease H domain of HIV-1 reverse transcriptase. Science 1991;252:88–95. [PubMed: 1707186]

43. Chattopadhyay D, Finzel BC, Munson SH, Evans DB, Sharma SK, Strakalaitis NA, Brunner DP,Eckenrode FM, Dauter Z, Betzel C, Einspahr HM. Crystallographic analyses of an active HIV-1ribonuclease H domain show structural features that distinguish it from the inactive form. ActaCrystallogr. Sect 1993;49:423–427.

44. Pari K, Mueller GA, DeRose EF, Kirby TW, London RE. Solution structure of the RNase H domainof the HIV-1 reverse transcriptase in the presence of magnesium. Biochemistry 2003;42:639–650.[PubMed: 12534276]

45. Mueller GA, Pari K, DeRose EF, Kirby TW, London RE. Backbone dynamics of the RNase H domainof HIV-1 reverse transcriptase. Biochemistry 2004;43:9332–9342. [PubMed: 15260476]

46. Jacobo-Molina A, Ding J, Nanni RG, Clark AD, Lu X, Tantillo C, Williams RL, Kamer G, FerrisAL, Clark P, Hizi A, Hughes SH, Arnold E. Crystal structure of human immunodeficiency virus type1 reverse transcriptase complexed with double-stranded DNA at 3.0 Å resolution shows bent DNA.Proc. Natl. Acad. Sci. U.S.A 1993;90:6320–6324. [PubMed: 7687065]

47. Bebenek K, Berad WA, Darden TA, Prasad R, Luxon BA, Gorenstein DG, Wilson SM, Kunkel TA.A minor groove binding track in reverse transcriptase. Nat. Str. Biol 1997;4:194–197.

Somasunderam et al. Page 11

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 1.The DNA library sequence. The random region is marked as N22. The sequence has a 19-merforward primer and a 21-mer reverse primer.

Somasunderam et al. Page 12

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 2.Sequences of the selected thioaptamers after the 12th and 14th rounds of selection. Only thevariable region of the thioaptamer is shown. Sequences showing higher degrees of alignmentare grouped in class I and those with lower degrees of alignment are in class II. Truncatedsequences are listed under class III.

Somasunderam et al. Page 13

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 3.Binding of thioaptamer R12-2 to different proteins. EMSA of R12-2 shows binding to theRNase H domain of HIV-1 RT protein (■) but not to the E.coli RNase H (●). Initial libraryclone (▲) did not show binding to the HIV-1 RT. The intensity of the free-DNA band is plottedagainst the protein concentrations.

Somasunderam et al. Page 14

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 4.EMSA of the lead thioaptamer R12-2 to HIV-1 RT. (A) Gel electrophoresis mobility shiftassay showing binding of R12-2 to the HIV-1 RT. The arrows show the positions of upper andlower bands. Lane 1, no protein. Lanes 2-8 contain proteins in increasing concentrations (0.05,0.1, 0.2, 0.4, 0.6, 0.8 and 1μM, respectively). (B) Quantitative measurement of the intensity ofthe DNA band bound to the protein. Intensity of the upper band is plotted against the HIV RTconcentration in μM. (Bound band intensities are blanked against the intensity of the band atthe HIV RT concentration of 0.0)

Somasunderam et al. Page 15

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 5.Effect of the selected aptamer R12-2 on RNase H activity of HIV-1 RT. Radioactivity retainedin the filtrate, considered as a measure of RNase H activity, is plotted against the thioaptamerconcentration.

Somasunderam et al. Page 16

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 6.Effect of R-12-2 in cell infection. U373-R5 indicator cells were transfected with R-12-2 orrandom 62-mer and then infected the next day with SF-162 virus strain. Viral replication wasdetermined after 48h measuring β-gal activity in cell lysates by luminometry. Controls includemock transfected cells (OF only), thioaptamers without transfection reagent and AZT 10 μMat the time of infection. Significant differences compared to the mock-transfected controls areindicated by asterisks.

Somasunderam et al. Page 17

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 7.Comparison of inhibition by thioaptamers R12-2 and XBY-S2. Viral replication wasdetermined after 48h measuring β-gal activity in cell lysates by luminometry. Controls includemock transfected cells (OF only), thioaptamers without transfection reagent and AZT 10 μMat the time of infection.

Somasunderam et al. Page 18

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 8.Dose response of thioaptamer R12-2 (●) and XBY-S2 (○) treatment. Indicator U373-MAGI-CCR5 cells were transfected with various doses of thioaptamer 24 hrs prior to infection withHIV SF162, and analyzed for β-gal activity (RLU) after 48 hrs. Results are presented as % ofvirus production compared with untreated controls.

Somasunderam et al. Page 19

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

FIGURE 9.Effect of thioaptamer R12-2 with increasing virus inocula. Two fold dilutions of SF162supernatant were used to infect U373-MAGI-CCR5 cells 24 hrs after thioaptamer transfectionor control condition, and infection was quantified in cell lysates after 48 hrs (RLU). Closedcircles (●) represent untransfected cells, closed squares (■) represent cells transfected withR12-2, and open triangles (△) represent control AZT (10 μM).

Somasunderam et al. Page 20

Biochemistry. Author manuscript; available in PMC 2008 September 8.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript